One morning in late 1997, Stanley Miller lifted a glass vial from a cold, bubbling vat. For 25 years he had tended the vial as though it were an exotic orchid, checking it daily, adding a few pellets of dry ice as needed to keep it at –108 degrees Fahrenheit. He had told hardly a soul about it. Now he set the frozen time capsule out to thaw, ending the experiment that had lasted more than one-third of his 68 years.
Miller had filled the vial in 1972 with a mixture of ammonia and cyanide, chemicals that scientists believe existed on early Earth and may have contributed to the rise of life. He had then cooled the mix to the temperature of Jupiter’s icy moon Europa—too cold, most scientists had assumed, for much of anything to happen. Miller disagreed. Examining the vial in his laboratory at the University of California at San Diego, he was about to see who was right.
As Miller and his former student Jeffrey Bada brushed the frost from the vial that morning, they could see that something had happened. The mixture of ammonia and cyanide, normally colorless, had deepened to amber, highlighting a web of cracks in the ice. Miller nodded calmly, but Bada exclaimed in shock. It was a color that both men knew well—the color of complex polymers made up of organic molecules. Tests later confirmed Miller's and Bada’s hunch. Over a quarter-century, the frozen ammonia-cyanide blend had coalesced into the molecules of life: nucleobases, the building blocks of RNA and DNA, and amino acids, the building blocks of proteins. The vial’s contents would support a new account of how life began on Earth and would arouse both surprise and skepticism around the world.
Although one of Miller’s final experiments, it certainly wasn’t the final word. The last several years have seen a steady stream of corroborating evidence, including one experiment—so new it has not yet been published—that Miller’s colleague, the late Leslie Orgel, called “astonishing.”
For decades, those studying the origin of life have imagined that it emerged in balmy conditions from primordial soups, tropical ponds, even boiling volcanic vents. Miller and a few other scientists began to suspect that life began not in warmth but in ice—at temperatures that few living things can now survive. The very laws of chemistry may have favored ice, says Bada, now at the Scripps Institution of Oceanography in La Jolla, California. “We’ve been arguing for a long time,” he says, “that cold conditions make much more sense, chemically, than warm conditions.”
Miller’s frozen experiment is a striking testament to the idea. Although life requires liquid water, small amounts of liquid can persist even at –60°F. Microscopic pockets of water within the ice may have gathered simple molecules like the ones Miller synthesized, assembling them into longer and longer chains. A single cubic yard of sea ice contains a million or more liquid compartments, microscopic test tubes that could have created unique mixtures of RNA that eventually formed the first life.
If life on Earth arose from ice, then our chances of finding life elsewhere in the solar system—not to mention elsewhere in the galaxy—may be better than we ever imagined.
The vial of ammonia and cyanide chilling in Miller’s lab was just one of the chemical cocktails he kept, aging like wine in a cellar. Some of the samples sat in freezers, others under the sink, and still others in water baths maintained at various temperatures. They were part of an effort to understand chemical reactions that must have unfolded over millennia on early Earth. The location of every sample was stored in Miller’s head; occasionally he would give one to a student to analyze.
Matthew Levy, once a graduate student of Miller’s and now a molecular biologist at the Albert Einstein College of Medicine in New York City, recalls being handed one of the 25-year-old samples to work on. “I was scared,” he says. “I was thinking, these samples are older than I am.” Levy burned holes in his shirts over the next few weeks as he dissolved the samples with hydrochloric acid and ran them through an instrument called a high-performance liquid chromatograph to identify the chemicals that had formed. Red and green pens on the device traced out telltale peaks on a scrolling strip of paper. Those peaks corresponded to seven different amino acids and 11 types of nucleobases.
“What was remarkable,” Bada says, “is that the yield in these frozen experiments was better, for some compounds, than it was with room-temperature experiments.”
There were people who found the results a little too remarkable. When Bada and Miller submitted their findings to a top-tier science journal, the article was rejected. A reviewer of the manuscript felt that those molecules must surely have formed while the samples were thawing, not while frozen at the ridiculously low temperature of –108°F. So Miller, Bada, and Levy did more experiments to show that thawing played no role. They published their results in another journal, Icarus, in 2000.
The skepticism they faced was understandable. Chemical reactions do slow down as the temperature drops, and according to standard calculations, the reactions that assemble cyanide molecules into amino acids and nucleobases should run a hundred thousand times more slowly at –112°F than at room temperature. By that reckoning, even if Miller had run his experiment for 250 years—let alone 25—he should have seen nothing.
This is the main argument against Miller’s experiment, and against a cold origin of life in general. But strange things happen when you freeze chemicals in ice. Some reactions slow down, but others actually speed up—especially reactions that involve joining small molecules into larger ones. This seeming paradox is caused by a process called eutectic freezing. As an ice crystal forms, it stays pure: Only molecules of water join the growing crystal, while impurities like salt or cyanide are excluded. These impurities become crowded in microscopic pockets of liquid within the ice, and this crowding causes the molecules to collide more often. Chemically speaking, it transforms a tepid seventh-grade school dance into a raging molecular mosh pit.
“Usually as you cool things, the reaction rates go down,” concluded Leslie Orgel, who studied the origins of life at the Salk Institute in La Jolla, California, from the 1960s until his death last October. “But with eutectic freezing, the concentrations go up so fast that they more than make up” for the difference.
Cyanide is a good candidate as a precursor molecule in the life-in-a-freezer model for several reasons. First, planetary scientists suspect that cyanide was abundant on early Earth, deposited here by comets or created in the atmosphere by ultraviolet light or by lightning (once the atmosphere became oxygen rich, 2.5 billion years ago, the process would have stopped). Second, although cyanide is lethal to modern animals, it has a convenient tendency to self-assemble into larger molecules. Third, and perhaps most important, no matter how much cyanide rained down, it could become concentrated only in a cold environment—not in warm coastal lagoons—because it evaporates more quickly than water.
“The strong point of freezing,” according to Orgel, “is that you concentrate things very efficiently without evaporation.” Freezing also helps preserve fragile molecules like nucleobases, extending their lifetime from days to centuries and giving them time to accumulate and perhaps organize into something more interesting—like life.
Orgel and his coworkers proposed these ideas in 1966, when he showed that frozen cyanide efficiently assembles into larger molecules. Alan Schwartz, a biochemist at the University of Nijmegen in the Netherlands, took the idea further when he showed in 1982 that frozen cyanide, in the presence of ammonia, can form a nucleobase called adenine. And Stanley Miller likely had the eutectic effect in mind when he stowed his now famous samples in a freezing chamber full of dry ice and acetone.
While Miller and Orgel followed their clues in the lab, other scientists pursued their obsession with life’s chilly origins to the ends of the earth.
In July 2002 a small skiff dropped Hauke Trinks on the beach of Nordaustland, a rocky island encased in glaciers and nearly devoid of plants. Trinks, then a physicist at the Technical University of Hamburg-Harburg in Germany, had come to Nordaustland—far north of the Arctic Circle—to peer 4 billion years back in time to an era shortly after the end of the bombardment of Earth by asteroids. According to some solar evolution models, the sun was some 30 percent dimmer at that time, providing less heat to Earth. So as soon as the hail of asteroids stopped, Earth may have cooled to an average surface temperature of –40°F and a crust of ice as much as 1,000 feet thick may have covered the oceans. Many scientists have puzzled over how life could have arisen on a planet that was essentially a giant snowball. The answer, Trinks suspected, involved sea ice.
Trinks had become interested in sea ice 10 years before, while studying its tendency to accumulate pollutants from the atmosphere and concentrate them in liquid pockets within the ice. He set out to explore whether a layer of ice covering early Earth’s oceans might have gathered and assembled organic molecules.
With a few crates of supplies and two sled dogs, Trinks and his partner, Marie Tieche, hunkered down in a cabin on Nordaustland for 13 months. Each morning they monitored the temperature of the ice and prepared the day’s experiments. To study the networks of liquid pockets, Trinks injected dyes into the ice and watched through a microscope as they spread.
Winter deepened, 24-hour darkness descended, and the mercury plummeted to –20°F. Trinks continued his experiments, sometimes banging pans together to chase polar bears away. Once a walrus lunged up through the ice and dragged several of Trinks’s instruments into the ocean.
He built a makeshift lab table from planks of wood and discarded gasoline cans. He examined slices of sea ice under the microscope, his hood pulled tight around his eyes. Turning a knob with a gloved hand, he nudged a metal electrode nearly as fine as a red blood cell closer to an ice crystal. The needle on his voltmeter jerked sideways, registering a sharp drop in voltage on the crystal’s surface—evidence of a microscopic electric field that might arrange and orient molecules on the ice’s surface.
By the time Trinks returned to Hamburg in 2003, he had formulated a theory that ice was doing much more than just concentrating chemicals. The ice surface is a checkerboard of positive and negative charges; he imagined those charges grabbing individual nucleobases and stacking them like Pringles in a can, helping them coalesce into a chain of RNA. “The surface layer between ice and liquid is very complicated,” he says. “There is strong bonding between the surface of the ice and the liquid. Those bondings are important for producing long organic chains like RNA.”
At a lecture in Hamburg in 2003, Trinks met up with chemist Christof Biebricher, who was studying how the first RNA chains could have formed in the absence of the enzymes that guide their formation in living cells. Trinks approached Biebricher with his sea ice theory, but to Biebricher, the experiments to test it sounded messy—more like a margarita recipe than a serious scientific investigation. “Chemists,” says Biebricher, “do not like heterogeneous substances like ice.” But Trinks convinced him to try it in his laboratory at the Max Planck Institute for Biophysical Chemistry in Göttingen, Germany.
Biebricher sealed small amounts of RNA nucleobases—adenine, cytosine, guanine—with artificial seawater into thumb-size plastic tubes and froze them. After a year, he thawed the tubes and analyzed them for chains of RNA.
For decades researchers had tried to coax RNA chains to form under all sorts of conditions without using enzymes; the longest chain formed, which Orgel accomplished in 1982, consisted of about 40 nucleobases. So when Biebricher analyzed his own samples, he was amazed to see RNA molecules up to 400 bases long. In newer, unpublished experiments he says he has observed RNA molecules 700 bases long. Biebricher’s results are so fantastic that some colleagues have wondered whether accidental contamination played a role. Orgel defended the work. “It’s a remarkable result,” he said. “It’s so remarkable that everyone wants better evidence than they would for an unremarkable result. But I think it’s right.”
Biebricher had loaded the deck somewhat, because he wasn’t growing RNA chains from nothing. Before he froze his samples, he added an RNA template—a single-strand chain of RNA that guides the formation of a new strand of RNA. As that new RNA strand grows, it adheres to the template like one half of a zipper to the other. This must be how the first genes, made of RNA, would have copied themselves. But the first step was the formation of the original RNA molecule that served as a template, and how that step happened remains a mystery.
Ice may prove the crucial ingredient here, too. Deamer and his former student Pierre-Alain Monnard (now at Los Alamos National Laboratory in New Mexico) have run experiments frozen at 0°F for a month, without the aid of templates. In those relatively brief experiments they already see RNA molecules up to 30 bases long, at least as long as other researchers have seen in similar experiments without ice.
That is a good start, but it leaves unanswered the question: How do you get from tiny snippets of RNA to longer, well-crafted chains that could have acted as the first enzymes, doing fancy things like copying themselves The shortest RNA enzyme chains known today are about 50 bases long; most have more than 100. To work effectively, moreover, an RNA enzyme must fold correctly, which requires exactly the right sequence of bases.
A young scientist named Alexander Vlassov stumbled upon a possible answer. He was working at SomaGenics, a biotech company in Santa Cruz, California, to develop RNA enzymes that latch on to the hepatitis C virus. His RNA enzymes were behaving strangely: They normally consisted of a single segment of RNA, but every time he cooled them below freezing to purify them, the chain of RNA spontaneously joined its ends into a circle, like a snake biting its tail. As Vlassov worked to fix the technical glitch, he noticed that another RNA enzyme, called hairpin, also acted strangely. At room temperature, hairpin acts like scissors, snipping other RNA molecules into pieces. But when Vlassov froze it, it ran in reverse: It glued other RNA chains together end to end.
Vlassov and his coworkers, Sergei Kazakov and Brian Johnston, realized that the ice was driving both enzymes to work in reverse. Normally when an enzyme cuts an RNA chain in two, a water molecule is consumed in the process, and when two RNA chains are joined, a water molecule is expelled. By removing most of the liquid water, the ice creates conditions that allow the RNA enzyme to work in just one direction, joining RNA chains.
The SomaGenics scientists wondered whether an icy spot on early Earth could have driven a primitive enzyme to do the same. To investigate this, they introduced random mutations into the hairpin RNA, shortened it from its normal length of 58 bases, and even cut it into pieces—all in an effort to produce RNA enzymes that were as dodgy and imperfect as early Earth’s first enzymes likely were. These pseudoprimitive RNA enzymes do nothing at room temperature. But freeze them and they become active, joining other RNA molecules at a slow but measurable rate.
These findings inspired a theory that the first, extremely inefficient RNA enzymes got help from ice, which created an environment that encouraged short segments of RNA to stick together and behave as a single, larger RNA molecule. “Freezing stabilizes the complexes formed from multiple pieces of RNA,” concludes Kazakov. “So small pieces of RNA could be enzymes, not just large 50-base molecules.”
Equally telling, the pseudoprimitive RNA enzymes that Vlassov made grabbed and joined just about any other molecule. Enzymes on early Earth might have done the same, joining random segments of 5 or 10 RNA bases to form a variety of sequences.
All these processes would occur in microscopic pockets of liquid within the ice. “You have billions and billions of different possibilities,” Trinks says, “because you have billions of these small channels,” each like a microscopic test tube containing a unique RNA experiment. On the young Earth, pockets of liquid could have expanded into a network of channels that mixed their contents during freeze-thaw cycles, like day-night temperature changes in summer. In winter, the liquid pores would have contracted and become isolated again, returning to their separate experiments. With all the mixing, something special might eventually have formed: an RNA molecule that made rough copies of itself. And as Earth warmed, these molecules might have found a home in newly thawed seas or ponds, where something even more complex might have emerged—such as a cell-like membrane. “You have something that is multiplying itself, and you have variation that is inherited,” says Antonio Lazcano, a biology researcher and professor at the National Autonomous University of Mexico, in Mexico City?. “There you have the onset of Darwinian evolution. I’m willing to call that living.”
No one can really know if this is how life began. Other theories posit that mineral surfaces organized key molecules or volcanic sources synthesized amino acids. These theories need not be mutually exclusive. Glaciers on early Earth could have scooped up mineral dust; volcanoes could have rained ash onto nearby sea ice. Primordial ice “must have been full of impurities,” Lazcano says, “and those impurities must have had catalytic effects, enhancing the synthesis or destruction of some compounds.”
Shortly after Miller finished his 25-year experiment, he suffered a stroke that ended his career. His laboratory, with 40 years of samples, was emptied in 2002 to make way for a building renovation. Experiments that had run for years or decades were discarded without ever being analyzed. As Bada rescued a few items from his mentor’s freezer, safety personnel stood by in hazmat suits, sent by university officials concerned about rumors of toxic cyanide. Any sample that couldn’t be identified was incinerated. Miller was present for a few hours of this ordeal, struggling to find words to identify the vials that he had known so well.
Miller died on May 20, 2007, but the provocative theory he helped nurture lives on. In the latest twist, Miller’s ideas are influencing not just theories about life’s origin on Earth but also investigations about the potential for life elsewhere in the solar system. In fact, it was a dinner conversation with Bada regarding Jupiter’s moon Europa that prompted Miller to open his 25-year-old samples back in 1997. While most scientists were focusing on the possibility of life in Europa’s ocean, he and Bada had been talking about what biochemistry might happen in the 10-mile-thick layer of ice atop the ocean. Those speculations are more relevant than ever, with recent discoveries of geysers on Saturn’s icy moon Enceladus and elaborate organic molecules on Titan, another Saturnian moon. Recent studies show that Mars too has vast quantities of buried ice, especially at its poles.
If life arose in one of these frozen zones, it might still exist there. Although life as we know it requires liquid water, there are places where life survives well below freezing. In the microscopic veins that permeate Arctic ice, for example, the high concentration of salt can maintain traces of water in a liquid state down to –65°F. Bacteria and diatoms inhabit those liquid veins, and Hajo Eicken, a glaciologist at the University of Alaska at Fairbanks, suspects that similar habitats could exist in the lower, warmer layers of ice on Europa, and perhaps on the other moons as well. “There’s potentially hundreds of meters of ice, if not maybe a few kilometers, that may well be quite habitable,” Eicken says.
Liquid water—and life—occurs in other cold places, too. Films of liquid water persist far below freezing, like coatings of condensation, on the surfaces of some minerals. Under some conditions, these films may stay liquid down to –90°F. Bacteria beneath films of liquid water only several molecules thick have been found clinging to microscopic grains of clay in ice cores from Greenland. Slowly consuming the iron in a single grain, these bacteria could get by for a million years before exhausting their food supply; at colder temperatures, where metabolic demands are lower, they might survive hundreds of millions of years.
If life arose in ice on Earth, then why not on Mars, Europa, or Enceladus? “You’ve got to keep an open mind in this business,” Bada says. “If I were going to make a bet about what we’d find if we discover life elsewhere in the universe, I would suspect it would be more cold-adapted than hot-adapted.”